FINAL EMPIRICAL PROJECT

BEE2041710032265

HIV/AIDS

Data Unveiled

World Wide Data

Data Web Scraped from:

https://en.wikipedia.org/wiki/HIV/AIDS_in_Africa

HIV/AIDS remains a pressing global health concern, with profound implications for individuals

and communities worldwide. The provided graph on above illustrates the percentage of adults

aged 15-49 living with HIV across different regions. Notably, Sub-Saharan Africa emerges as the

epicentre of the epidemic, demonstrating the highest prevalence rates by a large margin.

Additionally, regions such as the Caribbean and Eastern Europe, along with Central Asia,

exhibit significant burdens of the disease.

In this data-driven blog post, we delve into the prevalence of HIV/AIDS worldwide, shedding light on

its significance and the imperative to disseminate information on this crucial topic. By examining

where the disease is most prevalent globally, we gain insights into the scope of the issue and

potential challenges associated with data interpretation.

To collect data for the visualisations on this page, I employed web scraping techniques to

extract information from the relevant Articles and Wikipedia pages. Using a popular

coding language called Python, I created scripts to ‘scrape’ or ‘read’ the web pages and

store the data in tables to allow me to analyse the underlining trends. After cleaning the

data (checking for missing values and removing unwanted symbols), I was able to create

the visualisations you see throughout this whole blog post.

Collecting the Data

Sub-Saharan Data

Exploring further into Sub-Saharan Africa, we uncover alarming statistics that shows the

severity of the HIV/AIDS crisis within this region. Countries such as Eswatini, Lesotho, and

Botswana bear the highest occurrence of the disease, with prevalence rates ranging from

around 22% to nearly 30%.

Investigating the factors contributing to this disproportionate impact is vital, to understand

what trends within a country can lead to these infection levels. We will take a deeper look into

the distributions and relationships surrounding HIV/AIDS in Southern Africa later on in this

blog.

Data Web Scraped from:

https://en.wikipedia.org/wiki/List_of_countries_by_HIV/AIDS_adult_prevalence_rate

Economic Factors in

Southern Sub-Saharan Africa

To ensure the data was accurate for the tasks at hand I

compared the list of countries missing from the dataset

to a list of Sub-Saharan Countries and found that there

were only 3 missing entries. ‘Somalia’ is the largest on

these and most likely to cause miss leading trends

however, like the other two, it is not near the countries I

intend to look at for the rest of the blog and should not

have a significant impact.

In the visualisation above, a physical map of Southern Africa highlights the distribution of

HIV/AIDS prevalence among countries, with darker shades indicating higher percentages of

adults affected by the disease. A striking observation emerges as we notice the geographical

proximity of Eswatini and Lesotho, both situated within South Africa, yet exhibiting stark

differences in infection rates. This phenomenon prompts further investigation into the

underlying factors contributing to such disparities within neighbouring regions. Additionally, a

discernible trend emerges as we move northward into Africa, with countries showing

progressively lower prevalence rates. However, it's crucial to acknowledge the potential

influence of data availability and research efforts on these observed patterns. Particularly in

more remote areas, where research funding may be more limited.

Data Web Scraped from:

https://en.wikipedia.org/wiki/List_of_countries_by_HIV/AIDS_adult_prevalence_rate

Data Web Scraped from:

https://en.wikipedia.org/wiki/List_of_African_countries_by_GDP_(PPP)

Mapping HIV/AIDS in Africa

Mapping GDP(PPP) in Africa

Shifting our focus to the new visualisation above, the same map is presented, but this time,

darker shades indicate countries with lower Gross Domestic Product (GDP) based on

Purchasing Power Parity (PPP). GDP (PPP) represents the total value of goods and services

produced within a country adjusted for differences in price levels compared to other

countries, providing a more accurate comparison of economic performance. The analysis of

GDP (PPP) in relation to HIV/AIDS infections reveals intriguing trends.

The two maps show that the countries with lower GDP (PPP) exhibit much higher prevalence of

HIV/AIDs. This seems to be irrespective of their neighbouring countries and their own infection

rates, reflecting the intricate interplay between economic factors and health outcomes. This

correlation emphasises the importance of socioeconomic factors in shaping the prevalence

and impact of HIV/AIDS within communities and nations.

A Closer Look at Lesotho, Eswatini,

and South Africa

Analysing the map plots, particularly focusing on Lesotho, Eswatini, and South Africa,

reveals intriguing patterns worth exploring further. The plot above depicts the trends in

HIV prevalence from 1990 to 2020 sourced from the World Data Bank.

Distinctly, we see a peak in HIV incidences in both Eswatini and Lesotho in 1996 and a

peak in South Africa in 2000. Following these spikes there is a general downwards trend

in cases as healthcare innovations improved. Moreover, we observe the impact that GDP

(PPP) has on HIV incidences. This graph consolidates the link between the two, with

countries with higher GDP (PPP) consistently showing lower HIV cases.

Data Collected from:

https://ourworldindata.org/hiv-aids

A striking observation is the delay in the manifestation of the deadly symptoms of

HIV/AIDS, particularly evident in economically poorer countries. This delay is also

reflected in the onset of AIDS-related complications, suggesting a lag between HIV

infection and the development of severe health consequences. Specifically, our analysis

unveils a notable spike in HIV incidences around 1996, followed by a corresponding surge

in deaths in 2004, particularly pronounced in Eswatini and Lesotho.

These findings align with the hypothesis put forth by the World Health Organization

(WHO), which suggests that individuals with HIV typically exhibit signs of HIV-related

illness within 5–10 years of infection. Our data and accompanying graphs provide

empirical evidence supporting this hypothesis, as evidenced by the peak in deaths

occurring precisely within the anticipated range of 5-10 years following the peak in

incidences. This temporal alignment underscores the predictable progression of the

HIV/AIDS epidemic and underscores the importance of timely interventions and

healthcare strategies to mitigate its impact.

By juxtaposing data from multiple sources and drawing correlations between HIV

incidence, mortality rates, and economic indicators, we gain a more comprehensive

understanding of the complex dynamics driving the spread and impact of HIV/AIDS in

affected regions. These insights are invaluable for informing targeted interventions and

policies aimed at addressing the multifaceted challenges posed by the epidemic.

Data Collected from:

https://ourworldindata.org/hiv-aids

Contributing Factors to

AIDS Diagnoses

Data Collected from:

https://data.world/login?next=%2Fcity-of-ny%2Ffju2

dad%2Fworkspace%2Ffile%3Ffilename%3Ddohmh-hiv-aids-annual-report-1.csv

Displayed here is a variable importance plot revealing the key attributes influencing AIDS

diagnoses.

For this section I decided to utilise a Random Forest as I have some prior experience

using this method. To sum up quickly, the code takes lots of differing variables and is able

to rank them on how much of an effect they have on a desired outcome. For example in

this case we are looking at factors that lead to an Aids Diagnosis.

This data originates from New York state, acknowledging the potential presence of

inherent biases. However, it remains the most extensively funded and reliable publicly

accessible dataset at our disposal. Gender emerges as the predominant factor, exerting

the most significant influence on AIDS diagnoses, followed closely by race. These are

incredibly useful attribute to be aware of as they allow for governments to target their

funding and help directly to the ones who need it most.

Notably, the United Hospital Fund (UHF), intricately linked with borough demographics,

exhibits a substantial impact on the likelihood of an AIDS diagnosis. It's intriguing to note

that age appears to have the least influence on AIDS diagnoses according to this analysis.

Further investigation is imperative to uncover nuanced trends and underlying dynamics

within this observation. This insight prompts deeper exploration into the intricate

interplay of demographics and health outcomes, shedding light on critical aspects of

AIDS epidemiology and potentially informing targeted interventions and policies.

Deeper Analysis

In this series of three stratified graphs, each delineating distinct demographic segments

pertinent to AIDS diagnoses, we embark on an analytical journey elucidating critical

insights into the disease's epidemiology.

The first graph unveils a notable surge in AIDS diagnoses within the Asian/Pacific Islander

demographic - a noteworthy finding warranting deeper investigation to discern the

underlying drivers of this pronounced increase.

Transitioning to the second graph, a discernible skew towards male individuals in AIDS

diagnoses is clear, followed by transgender individuals—a pattern that resonates with

established epidemiological expectations. Notably, these trends maintain their

significance despite the absence of a substantial portion of age-related data.

Data Collected from:

https://data.world/login?next=%2Fcity-of-ny%2Ffju2

dad%2Fworkspace%2Ffile%3Ffilename%3Ddohmh-hiv-aids-annual-report-1.csv

Data Collected from:

https://data.world/login?next=%2Fcity-of-ny%2Ffju2

dad%2Fworkspace%2Ffile%3Ffilename%3Ddohmh-hiv-aids-annual-report-1.csv

Subsequently, our exploration extends to the third graph, delving into the distribution of

AIDS diagnoses across various age cohorts. While the variable importance analysis

suggests a muted correlation between age and AIDS diagnosis, closer scrutiny reveals a

critical caveat: the prevalence of null data obfuscating substantive insights. Upon

mitigating this null data, a nuanced binomial bell curve appears, peaking within the 40–49

age range.

While the current dataset may exhibit limitations, these preliminary glimpses offer

valuable insights, hinting at broader trends awaiting illumination through more

exhaustive and robust data collection methodologies.

In conclusion, our journey through the data-driven exploration of HIV/AIDS has

illuminated critical facets of this global health crisis. From the macroscopic view of global

prevalence to the microscopic analysis of demographic influences, each insight has

contributed to a deeper understanding of the epidemic's complexities.

It is clear that HIV/AIDS remains a formidable challenge with far-reaching implications

for individuals, communities, and societies worldwide. Yet, amidst the sobering statistics

and daunting challenges, there is hope.

Hope lies in the power of data to inform targeted interventions and policies, in the

resilience of communities to unite against adversity, and in the dedication of healthcare

professionals and researchers to push the boundaries of knowledge and innovation.

Our journey doesn't stop here. It's a reminder to keep moving forward, to fight for fair

healthcare access, and to support those affected by HIV/AIDS.

Together, let's work to create a future where HIV/AIDS is no longer a global problem, but